The molecular machinery expressed by cells to coalesce membranes is relatively simple in the case of influenza virus, where hemagglutinin (HA) alone is sufficient for membrane fusion. The debate on how low pH-induced refolding of HA produces membrane fusion centers on the order in which HA merges the lipids of fusing membranes and opens a fusion pore to connect their aqueous contents. Either HA can first cause a proteinaceous ionic channel which then guides lipids to merge, or HA can first produce a local lipidic connection which then breaks to form a pore. It is unknown whether such a lipid connection without a fusion pore (hemifusion), which is achievable for a mutant HA, is a non-physiological intermediate or a sequential step in the fusion mediated by wild-type HA (wt-HA). To detect normally short-living intermediates, we reduced the density and mobility of activated HA, while simultaneously measuring lipid mixing and fusion pore conductance using a system of simultaneous electrophysiological recording and fluorescence video microscopy that we developed. Substantial improvements were made this year to the software controlling and analysing these experiments. Wt-HA did cause hemifusion under these suboptimal conditions. However, rather than an intermediate to fusion, this hemifusion was stable. If no pores were open prior to onset of lipid mixing, pores would not open at all, suggesting that while activated HA mediates local hemifusion, it simultaneously limits expansion of this hemifusion intermediate and impedes lipid diffusion till fusion pore formation.